Sex- and Disease-Specific Inflammasome Signatures in Circulating Blood Leukocytes of Patients with Abdominal Aortic Aneurysm Xiaoyu Wu,1,2 Sinan Cakmak,2 Markus Wortmann,2 Maani Hakimi,2,3 Jian Zhang,1 Dittmar Böckler,2 and Susanne Dihlmann2* 1
Department of Vascular & Thyroid Surgery, The First Hospital of China Medical University, Shenyang, China; 2University Hospital Heidelberg, Department of Vascular and Endovascular Surgery, Heidelberg, Germany; and 3Vaskuläre Biomaterialbank Heidelberg (VBBH), Heidelberg, Germany
Male sex is a risk factor for abdominal aortic aneurysm (AAA). Within the AAA adventitia, infiltrating leukocytes express high levels of inflammasome components. To further elucidate the role of inflammatory cells in the pathogenesis of AAA, we here addressed expression and functionality of inflammasome components in peripheral blood mononuclear cells (PBMC) of AAA patients in association with sex. PBMC and plasma were isolated from 100 vascular patients, including 34 pairs of AAA patients and age/sexmatched non-AAA patients. Male PBMC were found to express significantly higher mRNA levels of AIM2, NLRP3, ASC (PYCARD), CASP1, CASP5, and IL1B (all P < 0.0001) than female PBMC. Within the male patients, PBMC of AAA patients displayed increased mRNA levels of NLRP3 (P = 0.044), CASP1 (P = 0.032) and IL1B (P = 0.0004) compared with matched non-AAA PBMC, whereas there was no difference between female AAA and non-AAA patients. The relative protein level of NLRP3 was significantly lower in PBMC lysates from all AAA patients than in matched controls (P = 0.038), whereas AIM2 and active Caspase-1 (p10) protein levels were significantly increased (P = 0.014 and P = 0.049). ELISA revealed significantly increased IL-1α (mean = 6.34 versus 0.01 pg/mL) and IL-1β plasma levels (mean = 12.07 versus 0.04 pg/mL) in AAA patients. The data indicate that male PBMC display a systemic proinflammatory state with primed inflammasomes that may contribute to AAA-pathogenesis. The AAA-specific inflammasome activation pattern suggests differential regulation of the sensors AIM2 and NLRP3 in inflammatory cells of AAA patients. Online address: http://www.molmed.org doi: 10.2119/molmed.2016.00035
INTRODUCTION Dilatation and potential rupture of abdominal aortic aneurysm (AAA) result in high morbidity and mortality rates. Epidemiological studies identified age, hypertension, smoking and male sex as major risk factors for AAA. Currently, surgical or endovascular repair is the major therapeutic option to prevent life-threatening disease progression. Despite increasing knowledge
on histopathological features of AAA from human tissue and animal studies, potential targets for prediction of the individual rupture risk and for efficient pharmaceutical interventions to attenuate expansion remain to be identified (1,2). A driving force in the pathogenesis of AAA is chronic inflammation, resulting in progressive remodeling and deterioration of the aortic wall (3). Recent findings on vascular cell biology have
Address correspondence to Susanne Dihlmann, Department of Vascular and Endovascular Surgery, University of Heidelberg, Im Neuenheimer Feld 110, 69120, Heidelberg, Germany. Phone: + + 49-6221-56-6494; Fax: + + 49-6221-56-7654; E-mail: [email protected]. Submitted February 2, 2016; Accepted for Publication July 11, 2016; Published Online (www.molmed.org) July 29, 2016.
508 | Wu ET AL. | MOL MED 22:508-518, 2016
elucidated the role of innate and adaptive immune cells in AAA pathogenesis (4). Among the immune cells, which accumulate in the aneurysm wall, are T-cells, B-cells, macrophages (3), mast cells (5) and neutrophils (6). The most important mediators associated with AAA include matrix metalloproteinases and cytokines, which are released from different cell types within the aneurysm wall (4). Accordingly, locally increased messenger ribonucleid acid (mRNA) and protein levels of tumor necrosis factor α (TNF-α), interferon-γ (IFN-γ) and interleukin 1β (IL-1β) have been identified as the most prominent cytokines in AAA tissue samples (7). In addition, we have recently demonstrated elevated levels of inflammasome components in lymphocytic infiltrates of AAA (8). Assembly of these inflammasomes is in turn required for Caspase-1-dependent processing
RESEARCH ARTICLE
and activation of IL-1β from its inactive precursor form (9). We thus hypothesized that the inflammasome pattern in tissue-infiltrating and peripheral leukocytes of AAA patients might differ from that in individuals without AAA. The inflammasomes are a family of cytoplasmic multiprotein complexes for danger signal recognition that are induced by pathogens or cell debris (damage associated molecular patterns, (DAMPS)) (9,10). Expression levels of the inflammasome components are tightly regulated in different cell types and require a two-phase induction: a priming signal stimulates gene expression of inactive precursors before a second signal can stimulate assembly of the multiprotein complex consisting of sensor proteins, the adaptor protein apoptosis-associated speck-like protein with a caspase activation and recruitment domain (ASC) and inflammatory caspases (Caspase-1 or Caspase-5). Upon assembly, the caspases are enzymatically cleaved by the complex resulting in their active form, which then stimulates enzymatic cleavage of IL-1β and Interleukin-18 (IL-18) from inactive precursors (9). Depending on the initial sensor, several subfamilies are distinguished: The NOD-like receptors (NLR), including NLRP3 and others, act as sensors for intracellular damage associated signals such as cholesterol crystals, nanoparticles, reactive oxygen species and others. A second subfamily, including Absent in Melanoma 2 (AIM2) acts as sensors for intracellular dsDNA (9). The disease-specific regulation of inflammasome expression and activation in circulating blood mononuclear cells of AAA patients is currently unknown. Considering the locally increased level of inflammasomes and inflammasomepriming cytokines in AAA tissue, we investigated whether this pattern is reflected in PBMC of AAA patients. Because it is well accepted that AAA predominantly affects male patients, we additionally analyzed differences in inflammasome activities between male and female AAA and non-AAA individuals.
MATERIALS AND METHODS Blood Samples and Patient Characteristics Venous blood was taken from vascular patients on the day of their hospitalization, according to the standard operating procedures of the Vascular Biobank Heidelberg (VBBH). All patients gave their written informed consent to the study, which was approved by the ethical committee of the University of Heidelberg (S-301/2013 and S-412/2013). The blood was processed for further studies within 4 h after venipuncture. A total of 100 blood samples were consecutively included in the study (62 from male, and 38 from female patients) and analyzed for mRNA
expression. The reasons for hospitalization (type of vascular disease), additional previous or chronic diseases, as well as sociodemographic and clinical data of the patients were obtained from the hospital’s information system. Blood from a subset of 34 AAA patients (hospitalized for open or endovascular intervention of their AAA) was sex- and age-matched to 34 non-AAA patients (hospitalized for treatment of peripheral artery disease (PAD) or carotid artery stenosis, with no evidence for current aortic aneurysms). These 34 pairs were additionally investigated by Western blot and plasma analysis. Patients’ characteristics are shown in Table 1. All patients (AAA and controls) were examined for their aortic diameter by duplex sonography, prior
Table 1A. Patient characteristics according to sex. Male patients (N = 62)
Female patients (N = 38)
P-Valuea
67.50 ± 9.81 61.29%(38/62) 93.44%(57/61) 22.58%(14/62) 86.89%(53/61)
64.18 ± 12.19 55.88%(19/34) 91.18%(31/34) 20.59%(7/34) 76.47%(26/34)
0.138 0.667 0.698 1.000 0.254
43.55%(27/62) 32.26%(20/62) 16.13% (10/62)
18.42%(7/38) 21.05%(8/38) 42.11%(16/38)
0.016 0.259 0.005
8.06%(5/62)
18.42%(7/38)
0.203
Risk Factors Age (years ± SD) Smoking (No.) Hypertension (No.) Diabetes (No.) Statin use (No.) Type of vascular disease AAA (No.) Carotid stenosis (No.) PAD (No.) Other vascular diseases (limb vessel malformations, angina abdominalis) (No.)
Table 1B: Characteristics of patients with AAA and non-AAA (control).
Risk Factors Age (years) ± SD Smoking (No.) Hypertension (No.) Diabetes (No.) Statin use (No.) Cholesterol level (mg/dl) ± SD Inflammmatory parameters CRP (mg/dl) ± SD Leucocytes (x103/μL) ± SD Max aortic diameter (cm) ± SD
AAA (N = 34)
Control (N = 34)
P-Valuea
68.76 ± 7.36 61.76% (21/34) 97.06% (33/34) 11.76% (4/34) 85.29% (29/34) 175.61 ± 32.88
69.44 ± 7.00 61.76% (21/34) 94.12% (32/34) 35.29% (12/34) 79.41% (27/34) 164.84 ± 48.18
0.699 1.000 1.000 0.043 0.752 0.308
13.97 ± 24.61 7.79 ± 1.83
17.03 ± 29.89 9.08 ± 3.18
0.660 0.049
5.81 ± 0.95
1.96 ± 0.57
< 0.00001
a
Data was analyzed using Student t test or Chi square test.
MOL MED 22:508-518, 2016 | Wu ET AL. | 509
ACTIVITY IN PBMC OF AAA PATIENTS
to their hospitalization for treatment. All AAA patients and 10 of the control patients were additionally examined by preoperative multislice computed tomographic angiography. For all patients (AAA and controls), the timeframe between their latest aortic imaging and venipuncture for the study was less than six months, generally within one month. Within the control group, three patients had undergone a previous aortic repair due to aneurysm dating back more than five years. Blood Separation Separation of plasma and PBMC from blood was performed as described previously (11). After separation, half of each PBMC pellet was lysed in RIPA buffer (50 mmol/L Tris-HCl, pH 7.5, 150 mmol/L NaCl, 0.5% sodium deoxycholate, 0.1% SDS, 0.5% NP-40, complemented with protease inhibitor Mix G, Serva) for protein analysis. The second half was lysed in RLT-buffer (RNeasy mini Kit, Qiagen) for RNA extraction. Lysates were frozen for cryopreservation at –80°C until further usage. RNA Extraction and qPCR Analysis Total RNA was extracted from cryopreserved PBMC pellets using RNeasy mini Kit (Qiagen) according to the manufacturer’s instructions. For expression analysis, 0.2 μg of the total RNA was reverse transcribed using oligo-dT primers and SuperScript III reverse transcriptase (Life Technologies) as described previously (8). The StepOnePlus Real time PCR System (Life Technologies) was used for mRNA level analysis of AIM2, NLRP3, ASC (PYCARD), CASP1, CASP5 and IL1B transcripts. Reagents and qPCR primers have been described previously (11). CT values were analyzed at a Ct threshold of 0.34. Relative mRNA expression of inflammasome genes was calculated by using individual standard curves and normalized to the mean values of three reference genes (GAPDH, B2M, and ACTB).
Western Blotting PBMC samples were lysed in RIPA buffer and homogenized by ultrasound. Twenty μg of each lysate were separated by SDS-PAGE an blotted onto nitrocellulose (0,45 μm for detection of AIM2, NLRP3, ASC; 0,2 μm for detection of Caspase-1 (p10 and p50). Paired samples were loaded side by side to ensure comparability. Detection of proteins was performed over night with primary antibodies diluted in appropriate buffers as recommended by the manufacturers. AIM2 (D5X7K) rabbit mAb, NLRP3 (D2P5E) rabbit mAb, cleaved caspase-1 (D57A2) rabbit mAb, and GAPDH (D16H11)XP rabbit mAb were purchased from Cell Signaling Technology, ASC rabbit pAb was obtained from Enzo Life Sciences, Caspase-1 p10 rabbit pAb was from Santa Cruz Biotechnology. Anti-rabbit IgG HRPlinked secondary antibody (Cell Signaling Technology) and Western-lightning plus ECL (Perkin Elmer) were used for detection by chemoluminescence. For relative quantification, signals of equal exposure times were densitometrically analyzed with ImageJ and normalized to corresponding GAPDH signals. ELISA Active, cleaved IL-1α (p18) and IL-1β (p17) in plasma from patients were quantified by the Duo-Set ELISA Development systems for human IL-1α or IL-1β, respectively (R&D Systems Europe) according to the recommendations of the manufacturer. Tissue Samples and Immunohistochemistry Formalin-fixed and paraffin-embedded biopsies derived from AAA patients were provided by the Vascular Biobank Heidelberg (VBBH) (approved by the ethical committee of the University of Heidelberg; S-412/2013). Processing and immunohistochemical staining was performed following standard procedures as described previously (8) by using the following antibodies: anti-AIM2 (Sigma
510 | Wu ET AL. | MOL MED 22:508-518, 2016
Prestige Antibodies; dilution 1:200), anti-NLRP3 (Cell Signaling Technology, dilution 1:500). Statistical Analysis All data was analyzed by IBM SPSS version 21 (IBM Corp). Figures were prepared using GraphPad Prism (GraphPad). Mann Whitney U test was performed to analyze the relative mRNA levels within the 100 vascular patients according to sex. To control for confounding, multivariate analysis was performed by logistic regression for each mRNA. The mRNA levels of male and female patients were adjusted for AAA and PAD, because these variables appeared to differ significantly between both sexes (Table 1A). Next, results from AAA patients were compared with age- and sex-matched controls (vascular patients without AAA) in a matched case-control study. The mRNA and protein levels of matched patients were analyzed by the Wilcoxon signed rank test for univariate analysis. The median, minimum and maximum were determined to be displayed in the figures and tables. To control for confounding, multivariate analysis was performed by conditional logistic regression. Matched pairs of individual mRNA levels were adjusted for history of smoking, hypertension, diabetes and hyperlipidaemia. Spearman′s rank correlation coefficients were determined to assess the linear association between mRNA levels and clinical parameters (such as cholesterol level, CRP, leukocyte count, maximal AAA-diameter). P values below 0.05 were interpreted as significant. All supplementary materials are available online at www.molmed.org. RESULTS Inflammasome-Related mRNA Levels in PBMC Are Significantly Higher in Male than in Female Vascular Patients Because women are known to experience lower rates of inflammatory diseases (12) and AAA incidence is
RESEARCH ARTICLE
a pproximately four times higher in men than in women, we initially compared the relative mRNA levels of AIM2, NLRP3, ASC, CASP1, CASP5 and IL1B in native PBMC of 100 consecutively hospitalized male (n = 62) and female (n = 38) patients with different vascular diseases. As shown in Table 2, the relative mRNA levels of all tested inflammasome components were significantly higher in PBMC from male than in PBMC from female patients (AIM2: 1.03 versus 0.73, P = 0.00009; NLRP3: 1.59 versus 0.58, P = 0.000038; ASC = 2.89 versus 1.24, P = 0.000004; CASP1: 1.06 versus 0.67, P = 0.000004; CASP5: 1.22 versus 0.18, P < 0.000001; IL1B: 2.34 versus 0.25, P < 0.000001). In contrast, cardiovascular risk factors such as age, smoking or hypertension did not differ between male and female patients (Table 1A). Moreover, statin use, which might interfere with systemic inflammation, was similar in both sexes (Table 1A). Univariate logisitc regression revealed highly significant odds ratio (OR) of the individual mRNA levels for male
sex (AIM2: OR = 5.66, P = 0.001; NLRP3: OR = 2.04, P = 0.005; ASC: OR = 1.79, P = 0.001; CASP1: OR = 8.75, P = 0.0003; CASP5: OR = 8.15, P = 0.000014; IL1B: OR = 1.54, P = 0.002). After adjustment for AAA and PAD as possible confounding diseases, all inflammasome mRNA levels in PBMC were still significantly associated with male sex (Table 3). Inflammasome-Related mRNA Levels in PBMC Differ between Male AAA and Non-AAA Control Patients As expected, among the consecutively enrolled vascular patients in this study, the rate of AAA patients was significantly higher (P = 0.016) in male (27/62) than in female (7/38) (Table 1A). To further explore inflammasome activities particularly in PBMC of AAA patients, the relative mRNA levels of AIM2, NLRP3, ASC, CASP1, CASP5 and IL1B derived from the subgroup of 34 AAA patients were analyzed separately and compared with 34 ageand sex-matched controls. CASP1 and
Table 2. Relative mRNA levels of inflammasome components in PBMC normalized to ACTB/GAPDH/B2M. Data are derived from 62 male and 38 female patients. Target genes AIM2 NLRP3 ASC (PYCARD) CASP1 CASP5 IL1B
Male Median [range]
Female Median [range]
P- Valuea
1.03 [0.14–9.42] 1.59 [0.42–9.90] 2.89 [0.75–13.74] 1.06 [0.41–4.47] 1.22 [0.00–5.58] 2.34 [0.07–45.06]
0.73 [0.04–1.72] 0.58 [0.01–4.88] 1.24 [0.39–6.56] 0.67 [0.19–2.25] 0.18 [0.01–2.01] 0.25 [0.00–10.67]
0.00009 0.000038 0.000004 0.000004 < 0.000001 < 0.000001
a
Data was analyzed using the Mann Whitney U test.
Table 3. Logistic regression to identify independent variables associated with male sex. With adjustmenta
Without adjustment Target Genes
Odds Ratio
95% CI
P- Value
Odds Ratio
95% CI
P- Value
AIM2 NLRP3 ASC CASP1 CASP5 IL1B
5.66 2.04 1.79 8.75 8.15 1.54
2.12–15.09 1.24–3.35 1.27–2.52 2.67–28.68 3.16–21.04 1.18–2.03
0.001 0.005 0.001 0.0003 0.000014 0.002
4.81 1.78 1.71 7.21 8.66 1.51
1.79–12.90 1.06–3.01 1.20–2.42 2.14–24.29 3.03–24.79 1.13–2.01
0.002 0.030 0.003 0.001 0.00006 0.005
a
Each gene was individually adjusted for AAA and PAD.
IL1B mRNA levels were again found to be significantly increased in PBMC of AAA patients (CASP1: P = 0.028; IL1B: P = 0.0001, both sexes). In contrast, the relative mRNA levels of AIM2, NLRP3, ASC, and CASP5 did not differ between PBMC from AAA patients and controls (Supplementary Figure S1 and Supplementary Table S1). When male and female AAA patients were analyzed separately, none of the inflammasome mRNAs were found to differ between female AAA and female non-AAA patients (Figure 1 A–F and Supplementary Table S2). In contrast, CASP1 and IL1B mRNA expressions were again found to be significantly increased in PBMC from male AAA patients when compared with matched PBMC from male non-AAA patients (medianCASP1 = 1.21 versus 1.11, P = 0.032; medianIL1B = 4.65 versus 1.56, P = 0.0004, Figure 1D and F). In addition, NLRP3 mRNA was significantly increased in PBMC from male AAA patients (medianNLRP3 = 1.96 versus1.54; P = 0.044, Figure 1B and Supplementary Table 2). Finally, male AAA patients showed a higher mRNA expression of AIM2, ASC, CASP1, and CASP5 than female AAA patients in their PBMC (each P < 0.05; Figure 1). We next analyzed whether the mRNA levels in male and female AAA patients are associated with clinical parameters (cholesterol level, leukocyte count, CRP, AAA-diameter). No correlations were found between any of the inflammasome mRNA levels and disease parameters in male patients. In contrast, the CASP5 mRNA level in PBMC of female patients was found to correlate positively (r = 0.775; P = 0.041) with the maximal AAA diameter (Figure 2). mRNA Levels of Inflammasome Components Correlate with Each Other in PBMC of AAA Patients Within the AAA patients, Spearman′s correlation test revealed a strong positive correlation of the inflammasome mRNA levels in PBMC with
MOL MED 22:508-518, 2016 | Wu ET AL. | 511
Figure 1. mRNA levels of inflammasome-related genes in PBMC of male and female AAA patients (M-AAA, F-AAA) versus non-AAA patients (M-control, F-control). Relative mRNA level of AIM2 (A), NLRP3 (B), ASC (C), CASP1 (D), CASP5 (E) and IL1B (F) in PBMC from vascular patients was analyzed by qPCR. CT values were calculated by using individual standard curves and normalized to the mean CT values of three reference genes (GAPDH, B2M, ACTB). P values are derived from Wilcoxon signed rank test.
ACTIVITY IN PBMC OF AAA PATIENTS
512 | Wu ET AL. | MOL MED 22:508-518, 2016
RESEARCH ARTICLE
Table 4. Conditional logistic regression to identify independent risk factors for AAA. With adjustmenta
Without adjustment Target Genes
Odds Ratio
95% CI
P- Value
Odds Ratio
95% CI
P- Value
AIM2 NLRP3 ASC CASP1 CASP5 IL1B
0.71 1.28 1.14 2.02 0.88 1.95
0.35–1.47 0.87–1.90 0.90–1.45 0.78–5.24 0.51–1.50 1.15–3.32
0.358 0.214 0.292 0.150 0.629 0.014
0.99 1.53 0.58 40.85 0.14 2.91
0.43–2.31 0.19–12.65 0.10–3.25 0.12– > 999.99 0.003–5.41 1.01–8.39
0.981 0.693 0.534 0.213 0.289 0.048
a
Figure 2. Correlation of CASP5 mRNA expression in PBMC of female AAA-patients. Spearman′s rank test was used for analysis.
each other (Supplementary Table S3): The AIM2 mRNA level correlated positively with ASC (r = 0.664; P
Department of Vascular & Thyroid Surgery, The First Hospital of China Medical University, Shenyang, China; 2University Hospital Heidelberg, Department of Vascular and Endovascular Surgery, Heidelberg, Germany; and 3Vaskuläre Biomaterialbank Heidelberg (VBBH), Heidelberg, Germany
Male sex is a risk factor for abdominal aortic aneurysm (AAA). Within the AAA adventitia, infiltrating leukocytes express high levels of inflammasome components. To further elucidate the role of inflammatory cells in the pathogenesis of AAA, we here addressed expression and functionality of inflammasome components in peripheral blood mononuclear cells (PBMC) of AAA patients in association with sex. PBMC and plasma were isolated from 100 vascular patients, including 34 pairs of AAA patients and age/sexmatched non-AAA patients. Male PBMC were found to express significantly higher mRNA levels of AIM2, NLRP3, ASC (PYCARD), CASP1, CASP5, and IL1B (all P < 0.0001) than female PBMC. Within the male patients, PBMC of AAA patients displayed increased mRNA levels of NLRP3 (P = 0.044), CASP1 (P = 0.032) and IL1B (P = 0.0004) compared with matched non-AAA PBMC, whereas there was no difference between female AAA and non-AAA patients. The relative protein level of NLRP3 was significantly lower in PBMC lysates from all AAA patients than in matched controls (P = 0.038), whereas AIM2 and active Caspase-1 (p10) protein levels were significantly increased (P = 0.014 and P = 0.049). ELISA revealed significantly increased IL-1α (mean = 6.34 versus 0.01 pg/mL) and IL-1β plasma levels (mean = 12.07 versus 0.04 pg/mL) in AAA patients. The data indicate that male PBMC display a systemic proinflammatory state with primed inflammasomes that may contribute to AAA-pathogenesis. The AAA-specific inflammasome activation pattern suggests differential regulation of the sensors AIM2 and NLRP3 in inflammatory cells of AAA patients. Online address: http://www.molmed.org doi: 10.2119/molmed.2016.00035
INTRODUCTION Dilatation and potential rupture of abdominal aortic aneurysm (AAA) result in high morbidity and mortality rates. Epidemiological studies identified age, hypertension, smoking and male sex as major risk factors for AAA. Currently, surgical or endovascular repair is the major therapeutic option to prevent life-threatening disease progression. Despite increasing knowledge
on histopathological features of AAA from human tissue and animal studies, potential targets for prediction of the individual rupture risk and for efficient pharmaceutical interventions to attenuate expansion remain to be identified (1,2). A driving force in the pathogenesis of AAA is chronic inflammation, resulting in progressive remodeling and deterioration of the aortic wall (3). Recent findings on vascular cell biology have
Address correspondence to Susanne Dihlmann, Department of Vascular and Endovascular Surgery, University of Heidelberg, Im Neuenheimer Feld 110, 69120, Heidelberg, Germany. Phone: + + 49-6221-56-6494; Fax: + + 49-6221-56-7654; E-mail: [email protected]. Submitted February 2, 2016; Accepted for Publication July 11, 2016; Published Online (www.molmed.org) July 29, 2016.
508 | Wu ET AL. | MOL MED 22:508-518, 2016
elucidated the role of innate and adaptive immune cells in AAA pathogenesis (4). Among the immune cells, which accumulate in the aneurysm wall, are T-cells, B-cells, macrophages (3), mast cells (5) and neutrophils (6). The most important mediators associated with AAA include matrix metalloproteinases and cytokines, which are released from different cell types within the aneurysm wall (4). Accordingly, locally increased messenger ribonucleid acid (mRNA) and protein levels of tumor necrosis factor α (TNF-α), interferon-γ (IFN-γ) and interleukin 1β (IL-1β) have been identified as the most prominent cytokines in AAA tissue samples (7). In addition, we have recently demonstrated elevated levels of inflammasome components in lymphocytic infiltrates of AAA (8). Assembly of these inflammasomes is in turn required for Caspase-1-dependent processing
RESEARCH ARTICLE
and activation of IL-1β from its inactive precursor form (9). We thus hypothesized that the inflammasome pattern in tissue-infiltrating and peripheral leukocytes of AAA patients might differ from that in individuals without AAA. The inflammasomes are a family of cytoplasmic multiprotein complexes for danger signal recognition that are induced by pathogens or cell debris (damage associated molecular patterns, (DAMPS)) (9,10). Expression levels of the inflammasome components are tightly regulated in different cell types and require a two-phase induction: a priming signal stimulates gene expression of inactive precursors before a second signal can stimulate assembly of the multiprotein complex consisting of sensor proteins, the adaptor protein apoptosis-associated speck-like protein with a caspase activation and recruitment domain (ASC) and inflammatory caspases (Caspase-1 or Caspase-5). Upon assembly, the caspases are enzymatically cleaved by the complex resulting in their active form, which then stimulates enzymatic cleavage of IL-1β and Interleukin-18 (IL-18) from inactive precursors (9). Depending on the initial sensor, several subfamilies are distinguished: The NOD-like receptors (NLR), including NLRP3 and others, act as sensors for intracellular damage associated signals such as cholesterol crystals, nanoparticles, reactive oxygen species and others. A second subfamily, including Absent in Melanoma 2 (AIM2) acts as sensors for intracellular dsDNA (9). The disease-specific regulation of inflammasome expression and activation in circulating blood mononuclear cells of AAA patients is currently unknown. Considering the locally increased level of inflammasomes and inflammasomepriming cytokines in AAA tissue, we investigated whether this pattern is reflected in PBMC of AAA patients. Because it is well accepted that AAA predominantly affects male patients, we additionally analyzed differences in inflammasome activities between male and female AAA and non-AAA individuals.
MATERIALS AND METHODS Blood Samples and Patient Characteristics Venous blood was taken from vascular patients on the day of their hospitalization, according to the standard operating procedures of the Vascular Biobank Heidelberg (VBBH). All patients gave their written informed consent to the study, which was approved by the ethical committee of the University of Heidelberg (S-301/2013 and S-412/2013). The blood was processed for further studies within 4 h after venipuncture. A total of 100 blood samples were consecutively included in the study (62 from male, and 38 from female patients) and analyzed for mRNA
expression. The reasons for hospitalization (type of vascular disease), additional previous or chronic diseases, as well as sociodemographic and clinical data of the patients were obtained from the hospital’s information system. Blood from a subset of 34 AAA patients (hospitalized for open or endovascular intervention of their AAA) was sex- and age-matched to 34 non-AAA patients (hospitalized for treatment of peripheral artery disease (PAD) or carotid artery stenosis, with no evidence for current aortic aneurysms). These 34 pairs were additionally investigated by Western blot and plasma analysis. Patients’ characteristics are shown in Table 1. All patients (AAA and controls) were examined for their aortic diameter by duplex sonography, prior
Table 1A. Patient characteristics according to sex. Male patients (N = 62)
Female patients (N = 38)
P-Valuea
67.50 ± 9.81 61.29%(38/62) 93.44%(57/61) 22.58%(14/62) 86.89%(53/61)
64.18 ± 12.19 55.88%(19/34) 91.18%(31/34) 20.59%(7/34) 76.47%(26/34)
0.138 0.667 0.698 1.000 0.254
43.55%(27/62) 32.26%(20/62) 16.13% (10/62)
18.42%(7/38) 21.05%(8/38) 42.11%(16/38)
0.016 0.259 0.005
8.06%(5/62)
18.42%(7/38)
0.203
Risk Factors Age (years ± SD) Smoking (No.) Hypertension (No.) Diabetes (No.) Statin use (No.) Type of vascular disease AAA (No.) Carotid stenosis (No.) PAD (No.) Other vascular diseases (limb vessel malformations, angina abdominalis) (No.)
Table 1B: Characteristics of patients with AAA and non-AAA (control).
Risk Factors Age (years) ± SD Smoking (No.) Hypertension (No.) Diabetes (No.) Statin use (No.) Cholesterol level (mg/dl) ± SD Inflammmatory parameters CRP (mg/dl) ± SD Leucocytes (x103/μL) ± SD Max aortic diameter (cm) ± SD
AAA (N = 34)
Control (N = 34)
P-Valuea
68.76 ± 7.36 61.76% (21/34) 97.06% (33/34) 11.76% (4/34) 85.29% (29/34) 175.61 ± 32.88
69.44 ± 7.00 61.76% (21/34) 94.12% (32/34) 35.29% (12/34) 79.41% (27/34) 164.84 ± 48.18
0.699 1.000 1.000 0.043 0.752 0.308
13.97 ± 24.61 7.79 ± 1.83
17.03 ± 29.89 9.08 ± 3.18
0.660 0.049
5.81 ± 0.95
1.96 ± 0.57
< 0.00001
a
Data was analyzed using Student t test or Chi square test.
MOL MED 22:508-518, 2016 | Wu ET AL. | 509
ACTIVITY IN PBMC OF AAA PATIENTS
to their hospitalization for treatment. All AAA patients and 10 of the control patients were additionally examined by preoperative multislice computed tomographic angiography. For all patients (AAA and controls), the timeframe between their latest aortic imaging and venipuncture for the study was less than six months, generally within one month. Within the control group, three patients had undergone a previous aortic repair due to aneurysm dating back more than five years. Blood Separation Separation of plasma and PBMC from blood was performed as described previously (11). After separation, half of each PBMC pellet was lysed in RIPA buffer (50 mmol/L Tris-HCl, pH 7.5, 150 mmol/L NaCl, 0.5% sodium deoxycholate, 0.1% SDS, 0.5% NP-40, complemented with protease inhibitor Mix G, Serva) for protein analysis. The second half was lysed in RLT-buffer (RNeasy mini Kit, Qiagen) for RNA extraction. Lysates were frozen for cryopreservation at –80°C until further usage. RNA Extraction and qPCR Analysis Total RNA was extracted from cryopreserved PBMC pellets using RNeasy mini Kit (Qiagen) according to the manufacturer’s instructions. For expression analysis, 0.2 μg of the total RNA was reverse transcribed using oligo-dT primers and SuperScript III reverse transcriptase (Life Technologies) as described previously (8). The StepOnePlus Real time PCR System (Life Technologies) was used for mRNA level analysis of AIM2, NLRP3, ASC (PYCARD), CASP1, CASP5 and IL1B transcripts. Reagents and qPCR primers have been described previously (11). CT values were analyzed at a Ct threshold of 0.34. Relative mRNA expression of inflammasome genes was calculated by using individual standard curves and normalized to the mean values of three reference genes (GAPDH, B2M, and ACTB).
Western Blotting PBMC samples were lysed in RIPA buffer and homogenized by ultrasound. Twenty μg of each lysate were separated by SDS-PAGE an blotted onto nitrocellulose (0,45 μm for detection of AIM2, NLRP3, ASC; 0,2 μm for detection of Caspase-1 (p10 and p50). Paired samples were loaded side by side to ensure comparability. Detection of proteins was performed over night with primary antibodies diluted in appropriate buffers as recommended by the manufacturers. AIM2 (D5X7K) rabbit mAb, NLRP3 (D2P5E) rabbit mAb, cleaved caspase-1 (D57A2) rabbit mAb, and GAPDH (D16H11)XP rabbit mAb were purchased from Cell Signaling Technology, ASC rabbit pAb was obtained from Enzo Life Sciences, Caspase-1 p10 rabbit pAb was from Santa Cruz Biotechnology. Anti-rabbit IgG HRPlinked secondary antibody (Cell Signaling Technology) and Western-lightning plus ECL (Perkin Elmer) were used for detection by chemoluminescence. For relative quantification, signals of equal exposure times were densitometrically analyzed with ImageJ and normalized to corresponding GAPDH signals. ELISA Active, cleaved IL-1α (p18) and IL-1β (p17) in plasma from patients were quantified by the Duo-Set ELISA Development systems for human IL-1α or IL-1β, respectively (R&D Systems Europe) according to the recommendations of the manufacturer. Tissue Samples and Immunohistochemistry Formalin-fixed and paraffin-embedded biopsies derived from AAA patients were provided by the Vascular Biobank Heidelberg (VBBH) (approved by the ethical committee of the University of Heidelberg; S-412/2013). Processing and immunohistochemical staining was performed following standard procedures as described previously (8) by using the following antibodies: anti-AIM2 (Sigma
510 | Wu ET AL. | MOL MED 22:508-518, 2016
Prestige Antibodies; dilution 1:200), anti-NLRP3 (Cell Signaling Technology, dilution 1:500). Statistical Analysis All data was analyzed by IBM SPSS version 21 (IBM Corp). Figures were prepared using GraphPad Prism (GraphPad). Mann Whitney U test was performed to analyze the relative mRNA levels within the 100 vascular patients according to sex. To control for confounding, multivariate analysis was performed by logistic regression for each mRNA. The mRNA levels of male and female patients were adjusted for AAA and PAD, because these variables appeared to differ significantly between both sexes (Table 1A). Next, results from AAA patients were compared with age- and sex-matched controls (vascular patients without AAA) in a matched case-control study. The mRNA and protein levels of matched patients were analyzed by the Wilcoxon signed rank test for univariate analysis. The median, minimum and maximum were determined to be displayed in the figures and tables. To control for confounding, multivariate analysis was performed by conditional logistic regression. Matched pairs of individual mRNA levels were adjusted for history of smoking, hypertension, diabetes and hyperlipidaemia. Spearman′s rank correlation coefficients were determined to assess the linear association between mRNA levels and clinical parameters (such as cholesterol level, CRP, leukocyte count, maximal AAA-diameter). P values below 0.05 were interpreted as significant. All supplementary materials are available online at www.molmed.org. RESULTS Inflammasome-Related mRNA Levels in PBMC Are Significantly Higher in Male than in Female Vascular Patients Because women are known to experience lower rates of inflammatory diseases (12) and AAA incidence is
RESEARCH ARTICLE
a pproximately four times higher in men than in women, we initially compared the relative mRNA levels of AIM2, NLRP3, ASC, CASP1, CASP5 and IL1B in native PBMC of 100 consecutively hospitalized male (n = 62) and female (n = 38) patients with different vascular diseases. As shown in Table 2, the relative mRNA levels of all tested inflammasome components were significantly higher in PBMC from male than in PBMC from female patients (AIM2: 1.03 versus 0.73, P = 0.00009; NLRP3: 1.59 versus 0.58, P = 0.000038; ASC = 2.89 versus 1.24, P = 0.000004; CASP1: 1.06 versus 0.67, P = 0.000004; CASP5: 1.22 versus 0.18, P < 0.000001; IL1B: 2.34 versus 0.25, P < 0.000001). In contrast, cardiovascular risk factors such as age, smoking or hypertension did not differ between male and female patients (Table 1A). Moreover, statin use, which might interfere with systemic inflammation, was similar in both sexes (Table 1A). Univariate logisitc regression revealed highly significant odds ratio (OR) of the individual mRNA levels for male
sex (AIM2: OR = 5.66, P = 0.001; NLRP3: OR = 2.04, P = 0.005; ASC: OR = 1.79, P = 0.001; CASP1: OR = 8.75, P = 0.0003; CASP5: OR = 8.15, P = 0.000014; IL1B: OR = 1.54, P = 0.002). After adjustment for AAA and PAD as possible confounding diseases, all inflammasome mRNA levels in PBMC were still significantly associated with male sex (Table 3). Inflammasome-Related mRNA Levels in PBMC Differ between Male AAA and Non-AAA Control Patients As expected, among the consecutively enrolled vascular patients in this study, the rate of AAA patients was significantly higher (P = 0.016) in male (27/62) than in female (7/38) (Table 1A). To further explore inflammasome activities particularly in PBMC of AAA patients, the relative mRNA levels of AIM2, NLRP3, ASC, CASP1, CASP5 and IL1B derived from the subgroup of 34 AAA patients were analyzed separately and compared with 34 ageand sex-matched controls. CASP1 and
Table 2. Relative mRNA levels of inflammasome components in PBMC normalized to ACTB/GAPDH/B2M. Data are derived from 62 male and 38 female patients. Target genes AIM2 NLRP3 ASC (PYCARD) CASP1 CASP5 IL1B
Male Median [range]
Female Median [range]
P- Valuea
1.03 [0.14–9.42] 1.59 [0.42–9.90] 2.89 [0.75–13.74] 1.06 [0.41–4.47] 1.22 [0.00–5.58] 2.34 [0.07–45.06]
0.73 [0.04–1.72] 0.58 [0.01–4.88] 1.24 [0.39–6.56] 0.67 [0.19–2.25] 0.18 [0.01–2.01] 0.25 [0.00–10.67]
0.00009 0.000038 0.000004 0.000004 < 0.000001 < 0.000001
a
Data was analyzed using the Mann Whitney U test.
Table 3. Logistic regression to identify independent variables associated with male sex. With adjustmenta
Without adjustment Target Genes
Odds Ratio
95% CI
P- Value
Odds Ratio
95% CI
P- Value
AIM2 NLRP3 ASC CASP1 CASP5 IL1B
5.66 2.04 1.79 8.75 8.15 1.54
2.12–15.09 1.24–3.35 1.27–2.52 2.67–28.68 3.16–21.04 1.18–2.03
0.001 0.005 0.001 0.0003 0.000014 0.002
4.81 1.78 1.71 7.21 8.66 1.51
1.79–12.90 1.06–3.01 1.20–2.42 2.14–24.29 3.03–24.79 1.13–2.01
0.002 0.030 0.003 0.001 0.00006 0.005
a
Each gene was individually adjusted for AAA and PAD.
IL1B mRNA levels were again found to be significantly increased in PBMC of AAA patients (CASP1: P = 0.028; IL1B: P = 0.0001, both sexes). In contrast, the relative mRNA levels of AIM2, NLRP3, ASC, and CASP5 did not differ between PBMC from AAA patients and controls (Supplementary Figure S1 and Supplementary Table S1). When male and female AAA patients were analyzed separately, none of the inflammasome mRNAs were found to differ between female AAA and female non-AAA patients (Figure 1 A–F and Supplementary Table S2). In contrast, CASP1 and IL1B mRNA expressions were again found to be significantly increased in PBMC from male AAA patients when compared with matched PBMC from male non-AAA patients (medianCASP1 = 1.21 versus 1.11, P = 0.032; medianIL1B = 4.65 versus 1.56, P = 0.0004, Figure 1D and F). In addition, NLRP3 mRNA was significantly increased in PBMC from male AAA patients (medianNLRP3 = 1.96 versus1.54; P = 0.044, Figure 1B and Supplementary Table 2). Finally, male AAA patients showed a higher mRNA expression of AIM2, ASC, CASP1, and CASP5 than female AAA patients in their PBMC (each P < 0.05; Figure 1). We next analyzed whether the mRNA levels in male and female AAA patients are associated with clinical parameters (cholesterol level, leukocyte count, CRP, AAA-diameter). No correlations were found between any of the inflammasome mRNA levels and disease parameters in male patients. In contrast, the CASP5 mRNA level in PBMC of female patients was found to correlate positively (r = 0.775; P = 0.041) with the maximal AAA diameter (Figure 2). mRNA Levels of Inflammasome Components Correlate with Each Other in PBMC of AAA Patients Within the AAA patients, Spearman′s correlation test revealed a strong positive correlation of the inflammasome mRNA levels in PBMC with
MOL MED 22:508-518, 2016 | Wu ET AL. | 511
Figure 1. mRNA levels of inflammasome-related genes in PBMC of male and female AAA patients (M-AAA, F-AAA) versus non-AAA patients (M-control, F-control). Relative mRNA level of AIM2 (A), NLRP3 (B), ASC (C), CASP1 (D), CASP5 (E) and IL1B (F) in PBMC from vascular patients was analyzed by qPCR. CT values were calculated by using individual standard curves and normalized to the mean CT values of three reference genes (GAPDH, B2M, ACTB). P values are derived from Wilcoxon signed rank test.
ACTIVITY IN PBMC OF AAA PATIENTS
512 | Wu ET AL. | MOL MED 22:508-518, 2016
RESEARCH ARTICLE
Table 4. Conditional logistic regression to identify independent risk factors for AAA. With adjustmenta
Without adjustment Target Genes
Odds Ratio
95% CI
P- Value
Odds Ratio
95% CI
P- Value
AIM2 NLRP3 ASC CASP1 CASP5 IL1B
0.71 1.28 1.14 2.02 0.88 1.95
0.35–1.47 0.87–1.90 0.90–1.45 0.78–5.24 0.51–1.50 1.15–3.32
0.358 0.214 0.292 0.150 0.629 0.014
0.99 1.53 0.58 40.85 0.14 2.91
0.43–2.31 0.19–12.65 0.10–3.25 0.12– > 999.99 0.003–5.41 1.01–8.39
0.981 0.693 0.534 0.213 0.289 0.048
a
Figure 2. Correlation of CASP5 mRNA expression in PBMC of female AAA-patients. Spearman′s rank test was used for analysis.
each other (Supplementary Table S3): The AIM2 mRNA level correlated positively with ASC (r = 0.664; P
